Diamond & Related Materials 15 (2006) 1601 – 1608
www.elsevier.com/locate/diamond
Thermionic emission from surface-terminated nanocrystalline diamond
Vance S. Robinson a , Yoshiyuki Show b , Greg M. Swain b ,
Ronald G. Reifenberger a , Timothy S. Fisher a,⁎
a
Birck Nanotechnology Center, Purdue University, 1205 W. State St., West Lafayette, IN 47907-2057, United States
b
Department of Chemistry Michigan State University East Lansing, MI 48824, United States
Received 1 July 2005; received in revised form 18 January 2006; accepted 22 January 2006
Available online 9 March 2006
Abstract
Thermionic electron emission forms the basis of both electron sources for a variety of applications and a direct energy conversion process that
is compact and scalable. The present study characterizes thermionic emission from boron-doped nanocrystalline diamond films with hydrogen and
nitrophenyl surface termination layers. A hemispherical energy analyzer was used to measure electron energy distributions from the emitters at
elevated temperatures. Thermionic emission energy distributions, acquired at temperatures ranging from 700 to 1100°C, reveal that emission
occurs from regions of differing work functions. The relative peak intensities, representing each work function, change with temperature indicating
instability in the emitter's surface chemistry. Corresponding partial pressure measurements of pertinent gases present in the chamber during the
experiment were collected by a residual gas analyzer and support the hypothesis of unstable surface chemistry. The lowest work functions
measured for the hydrogen- and nitrophenyl-terminated films were 3.95 and 3.88 eV, respectively. After the initial heating cycle, the hydrogenterminated sample's surface was regenerated by exposure to hydrogen plasma. The lower work function was restored by this process, and the
resulting thermionic electron energy distributions again were indicative of surface desorption.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Thermionic emission; Nanocrystalline diamond; Electron energy analyzer; Surface termination
1. Introduction
Thermionic emitters are used as electron sources in many
contemporary applications such as fluorescent lighting, cathode
ray tubes, X-ray tubes, mass spectrometers, vacuum gauges,
scanning electron microscopes, and other scientific instruments.
Most such sources are resistively heated filaments that produce a
flow of electrons. Further, thermionic emission provides a means
of converting heat directly into electrical power. Such thermionic converters have been designed to operate in conjunction
with various heat sources, such as solar radiation, nuclear reactions, and the combustion of fossil fuels [1–6]. Some of this
technology's more attractive qualities include: compactness,
scalability, and high waste heat rejection temperatures for cascaded systems [3,7]. Harnessing thermionic power generation
for efficient direct energy conversion is the underlying motivation of this work.
⁎ Corresponding author. Tel.: +1 765 494 5627.
E-mail address: tsfisher@purdue.edu (T.S. Fisher).
0925-9635/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.diamond.2006.01.017
In an effort to improve the thermodynamic efficiency and
capacity of electron emission, low-work-function materials must
be developed [8–10]. The properties of chemical vapor
deposited diamond with a hydrogen surface termination have
been investigated by several groups, who have demonstrated
that hydrogen-terminated boron-doped diamond has a negative
electron affinity [11–13], and a hydrogen-terminated nanocrystalline diamond (HND) sample is tested in this work. A nitrophenyl-terminated nanocrystalline diamond (NND) sample is
included in this study because nitrophenyl termination represents the successful modification of an otherwise non-reactive
surface with significant potential for electrochemical applications [14,15]. Surface temperature can affect the stability of such
termination layers, and the present work investigates the stability
and performance of hydrogen and nitrophenyl surface terminations on nanocrystalline diamond films at elevated temperatures.
Subsequent sections of this paper describe results from
thermionic emission testing of different nanocrystalline diamond films in detail. First, thermionic emission theory and its
relation to thermionic electron energy distributions (TEEDs) are
1602
V.S. Robinson et al. / Diamond & Related Materials 15 (2006) 1601–1608
described. This section also includes a discussion of instrument
effects on the reported measurements. Following the theory
section is a description of the experimental setup and the diamond film deposition process. TEED results and corresponding
theoretical curves are presented along with applicable partial
pressure results in the following section. A brief conclusion
follows, summarizing the results and the main contributions of
this work.
2. Theory
The thermionic emission energy distribution (TEED) from a
free-electron metal is given by [16]:
J VðEÞ ¼
4kmq
J3
ðE−/Þ
!
":
E
1 þ exp
kB T
ð1Þ
The terms m, q, and ℏ are the electron mass, electron charge and
reduced Planck's constant (h / 2π), respectively. The term E
represents electron energy; ϕ is the material's work function; kB
is Boltzman's constant; and T represents the temperature of the
emitter surface. The energy distribution is symbolized by J′
because it is the energy derivative of the saturation current
density, J. Differentiation of Eq. (1) with respect to electron
energy E reveals that the maximum thermionic emission intensity occurs at an energy that is kBT greater than the work
function. The foregoing relationship is used in subsequent sections to estimate work functions from the measured energies at
maximum intensity. Integration of Eq. (1) over all energies from
the ϕ to infinity gives the familiar Richardson–Dushman
equation [17]:
#
$
−/
J ¼ A⁎ T 2 exp
;
ð2Þ
kB T
where A⁎ represents the apparent emission constant [18].
In this study, a hemispherical energy analyzer was used to
measure TEEDs. The actual measured energy distribution is a
convolution of the distribution produced by the emitter and a
Gaussian instrument spreading function, which is determined
by specific analyzer parameters [19,20]. The form of the
instrument function is given by
" !
" #
1
1 E−E V 2
GI ¼ pffiffiffiffiffi exp −
:
ð3Þ
2
r
r 2k
The standard deviation σ is the term through which analyzer
parameters affect the function and is subsequently referred to as
the ‘analyzer resolution’. The convolution of Eqs. (1) and (3)
represents the energy distribution reported by the analyzer that
is subsequently used to quantify shifts in relative peak
intensities:
J Vobserved ¼
n
X
i¼1
Ai ½G⁎I Ii ð/i Þ&:
above the effective work function ϕi with an intensity Ii. The
summation accounts for the possibility of multiple regions on
the surface of the emitting material of differing effective work
functions ϕi.
In the present work, we interpret thermionic energy distributions through an effective mass approximation theory modified by
the Gaussian instrument function. The effective mass approximation is expected to be reasonable given that prior work by Köch
et al. [11] on thermionic field emission from polycrystalline
diamond has revealed relatively uniform surface emission
intensities explained by the thermal excitation of electrons into
the conduction band of a low-electron-affinity surface. We also
note that the effective mass model does not alter the shape of the
distribution predicted by free-electron theory, given that the mass
term is a pre-factor in the energy distribution [see Eq. (1)] that is
normalized in the presented results. These approximations are
also consistent, for example, with the use of Fowler–Nordheim
theory, which is based on free-electron theory, in interpretation of
field emission from polycrystalline diamond [21].
3. Experimental setup
Thermionic energy distributions were measured with a
hemispherical energy analyzer (SPECS-Phoibos 100 SCD)
connected to a vacuum chamber that achieves pressures of the
order of 10− 8 Torr. The heated emitter sample is located at the
focal plane, 40 mm below the analyzer's aperture. The temperature of the molybdenum heater (HeatWave Labs, Inc.) is
measured by a K-type thermocouple embedded 1mm below the
top surface and is modulated a PID-controlled power supply.
Due to contact resistance and other radiative heat losses between the surface of the sample and the thermocouple, the
temperature measurements can exceed the surface temperature
of the sample by as much as 30°C. All energy distributions
reported in this study were made after the temperature of the
emitter had stabilized for at least 20 min.
The heater assembly is thermally and electrically isolated
from the chamber by alumina hardware. The heater is negatively
biased (− 2 V) to accelerate electrons into the analyzer and to
to
analyzer
vacuum
chamber
dc
power
supply
Vbias
+
sample
heater
-
heater
+ controller
ð4Þ
where Ai is called the area coefficient because its magnitude is
proportional to the area of the sample that emits electrons at or
Fig. 1. Schematic of the experimental setup for measuring TEEDs including the
four main components: energy analyzer, heater, heater controller, and power
supply.
V.S. Robinson et al. / Diamond & Related Materials 15 (2006) 1601–1608
ensure that they have sufficient energy to overcome the work
function of the analyzer (4.12 eV). Electron acceleration was
achieved by connecting the heater to a DC power supply
(Hewlett Packard 6542A) and grounding the analyzer's aperture
(see Fig. 1). Voltage sense lines for the DC power supply were
implemented, reducing the uncertainty in the acceleration voltage to ± 0.3 mV. A residual gas analyzer (RGA, Inficon
Transpector 2) was used to measure changes in partial pressure
in the vacuum chamber that could represent changes in the
surface chemistry of the sample, and we note that operating the
RGA in conjunction with the analyzer caused a moderate
increase in instrument broadening.
4. Nanocrystalline diamond film growth
Nanocrystalline boron-doped diamond (BND) thin films
were deposited using microwave-assisted plasma enhanced
chemical vapor deposition (1.5 kW, 2.54 GHz, Astex, Inc.,
Lowell, MA) on highly conducting p-Si (100) substrates
(b0.001Ω-cm, Virginia Semiconductor, Inc.). The substrates
were pretreated by mechanical polishing on a felt pad using a
0.1 μm diameter diamond powder slurry in water (GE Superabrasives, Worthington, OH). The scratched substrates were
then ultrasonically cleaned in ultrapure water, isopropyl alcohol
(IPA), acetone, IPA, and ultrapure water to remove polishing
debris from the surface. Embedded diamond particles and polishing striations likely both serve as initial nucleation sites for
diamond growth. Increasing the nucleation density and nucleation rate of diamond growth on nondiamond substrates through
substrate pretreatment has been the subject of much research
over the years, and it is well established that scratching defects
and diamond particles function as nucleation sites for CVD
diamond growth [22]. The nanocrystalline diamond films were
deposited at 800 W, using an Ar/H2/CH4/B2H6 source gas mixture consisting of 1% CH4 / H2 + Ar with 10 ppm B2H6 added for
boron doping. The system pressure was 140 Torr, the substrate
temperature was ca. 800 °C (estimated via an optical pyrometer),
and the growth time was 2 h. All gases were ultrahigh purity
grade (99.999%, AGA Specialty Gas, Cleveland, OH). The
resulting film thickness was approximately 2μm.
Following the deposition, the CH4 and B2H6 flows were
stopped, and the films remained exposed to the H2/Ar plasma for
an additional 10min. The argon flow was then stopped, and the
plasma power and pressure were slowly reduced over a 5 min.
period to cool the samples in the presence of atomic hydrogen to
a temperature below 400 °C. The plasma power was then turned
off, and the films cooled to room temperature under a flow of H2.
This post-growth annealing in a hydrogen plasma serves to etch
away adventitious non-diamond carbon impurities, to minimize
dangling bonds, and to ensure maximum hydrogen termination.
The films had a nominal boron dopant concentration of
1020 B cm− 3 and a film resistivity of less than 0.05Ω-cm. Typical carrier concentrations (holes) are in the low 1020 cm− 3
range, and carrier mobilities are between 0.1 and 1 cm2/V-s, as
determined from Hall Effect measurements.
At this point, the fabrication of the hydrogen-terminated
(HND) film is complete. Further processing was required for the
1603
NND film. Boron-doped nanocrystalline diamond was chemically modified with nitrophenyl groups via the electroreduction
of 4-nitrophenyl diazonium salt [14,15,23–25]. The 1e−reduction reaction leads to the formation of a nitrophenyl radical at the
electrode surface, which then reacts with a surface atom and
attaches via a covalent bond. This electrochemically assisted
chemical modification scheme is a very versatile method for
controlling the surface chemistry of conductors and semiconductors with a wide variety of substituted aryl molecules. We
suppose that the covalent attachment of the nitrophenyl group
requires the formation of at least two nitrophenyl radicals. One
radical generates the “active” site on the hydrogen-terminated
diamond surface by abstracting a hydrogen atom. The second
nitrophenyl radical then couples at the newly formed radical site
on the surface. Fig. 2 contains an SEM image of the HND film in
which the sub-micrometer crystal facets are evident.
5. Results
Three TEEDs, corresponding to three temperature conditions, were recorded for each sample: at approximately 750 °C,
at a maximum temperature above 1090 °C, and again at 750 °C.
During the heating cycle, the residual gas analyzer monitored
partial pressures of relevant chemical species inside the chamber
that could represent desorption of the termination layer. The
following subsections describe the validation of the experiment
and thermionic emission results from the various samples.
5.1. Experiment validation
Before measuring thermionic emission from the nanocrystalline diamond films, emission from a material with known work
function was measured. The material was single-crystal tungsten (100), selected for its well documented work function in the
range of 4.52 to 4.59 eV [26–28]. Fig. 3 shows a thermionic
emission energy distribution from tungsten at approximately
900°C. The sharp increase in intensity followed by a gently
sloping high-energy tail is typical of thermionic emission [29].
Fig. 2. SEM image of the HND sample. Regions of differing work function are
likely for such heterogeneous surface. The length scale confirms nanoscale
crystal structure.
1604
V.S. Robinson et al. / Diamond & Related Materials 15 (2006) 1601–1608
Table 1
Resolution, energy of peak intensity, and effective work function of TEEDs
from the HND sample
TEED
Temperature
(°C)
Before max.
750
temp.
Max. temp.
1085
After max. temp. 750
Fig. 3. TEED from single-crystal tungsten at 900°C used to validate the
experimental setup.
Based on the curve fit to the data in Fig. 3, the work function of
the single-crystal tungsten sample is 4.56 eV, consistent with the
literature. These results for thermionic emission from tungsten
serve to validate the experimental setup used in this study.
5.2. Hydrogen-terminated nanocrytstalline diamond (HND)
TEEDs from the HND substrate deviate from a homogeneous
free-electron emitter as shown in Fig. 4, where the distributions
contain a secondary increase in intensity. Recalling Eq. (4),
additional peaks are interpreted as areas on the surface of the
emitter with different effective work functions. In Fig. 4, the
high-energy shoulder of the TEED at 750 °C before the
maximum temperature becomes the dominant peak at the
maximum temperature (1085 °C). Subsequently, the dominant
peak in the TEED before the maximum temperature reduces to
the leading shoulder seen in the TEED at 750 °C after the
maximum temperature. Table 1 summarizes the important
features of the three TEEDs for the HND sample, the locations
of their respective peaks, and the corresponding work functions.
Intensity (arb. units)
1
750C before max temp
1085C max temp
750C after max temp
0.8
0.6
0.4
0.2
0
3
3.5
4
4.5
5
5.5
6
6.5
Energy (eV)
Fig. 4. TEEDs from a hydrogen-terminated nanocrystalline diamond measured
at 750, 1085 and 750°C. The shift in the energy at maximum intensity before
and after the maximum temperature indicates that the relative area of higher
work function has increased in size.
Energy at max. Corresponding
intensity (eV) estimated
work function, ϕ (eV)
4.03
3.95
4.61
4.69
4.52
4.61
Considering that the dominant peak at and after the
maximum temperature corresponds to a work function very
similar to that of graphite 4.6eV [30], we believe that the high
temperature caused the termination layer to desorb such that the
majority of emitted electrons derive from regions on the surface
that contain π-bonded carbon atoms as a result of the loss of the
hydrogen surface termination. The nanocrystalline diamond
film also has a high fraction of exposed grain boundaries where
π-bonded carbon is known to exist [31,32].
The partial pressures in the vacuum chamber during the
heating cycle confirm the hypothesis that the surface termination is unstable at these temperatures. Fig. 5 shows a mass
spectrum based on partial pressures of molecules in the vacuum
chamber measured simultaneously with the TEEDs. A mass-tocharge ratio of 18 (corresponding to H2O) represents the largest
partial pressure for all conditions, and its presence is reduced
significantly by the heating cycle. The inset highlights three
particular charge-to-mass ratios, 1 (H), 2 (H2) and 16 (CH4) that
would likely be present upon desorption of atomic H or CH3
groups. According to the inset in Fig. 5, H2 occupied a slightly
greater percentage of the vacuum chamber gas after the heating
cycle than it did before. These results are contrary to the general
trend of diminishing partial pressures witnessed in the other
measured species. This response may indicate that H2 has been
supplied to the vacuum chamber from the hydrogen-terminated
film through a desorption process, but the measurement could
also be influenced of the decomposition of water into H2.
Recently, Köck et al. [9] published results of field emission
from surface-treated N-doped diamond films at elevated
temperatures. They observed a decrease in current at 725 °C
and attributed the decrease to the instability of the hydrogen
surface layer at temperatures above 725 °C. Furthermore, Hamza
et al. [33] observed that hydrogen on diamond desorbs at
temperatures between 800 and 1000 °C (∼1100–1300K) in
UHVand concluded that both surface-adsorbed and near-surface
absorbed hydrogen is released. The TEEDs measured in this
study support the hypothesis offered by Köck et al. Assuming
that the low-energy peak represents emission from a hydrogenterminated region, one can conclude that the hydrogen layer has
become unstable, and that electrons begin to experience a
generally higher work function at or slightly above 750°C.
5.3. Nanocrystalline diamond with nitrophenyl adlayer (NND)
Similar to the HND film, the nitrophenyl adlayer was unstable above 750 °C, and the maximum thermionic electron
V.S. Robinson et al. / Diamond & Related Materials 15 (2006) 1601–1608
0.12
750C before max temp
1085C max temp
2.50
750C after max temp
2.00
1.50
Partial Pressure, Nitrogen
Equivalent (x10-6 torr)
Partial Pressure, Nitrogen Equivalent
(x10-6 torr)
3.00
1605
0.10
0.08
0.06
0.04
0.02
0.00
1
2
16
(m/z) Charge-to-Mass Ratio
1.00
0.50
0.00
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Mass-to-Charge Ratio
Fig. 5. Mass spectra from hydrogen-terminated nanocrystalline diamond at 750, 1085 and again at 750°C. The inset displays the partial pressure of three particular
mass-to-charge ratios (1 and 2, and 16) corresponding to H, H2, and CH4, respectively. These species may form during desorption of the hydrogen termination layer.
intensity for this sample shifted to higher energy after heating to
the maximum temperature, as illustrated in Fig. 6. The initial
TEED at 750°C exhibits a similar peak followed by a higherenergy shoulder, which becomes the dominant peak at the
maximum temperature (1089 °C) and at 750°C after the maximum temperature. Comparing the data in Table 1 with that in
Table 2 below, it is clear that at the maximum temperature and
above, the energy distributions are dominated by emission from
a region with an effective work function in the range 4.53 to
4.61 eV. Also, before the maximum temperature, the NND film
exhibited the lowest recorded effective work function, 3.88 eV.
RGA data were recorded simultaneously with the TEEDs
during tests on the NND sample. The inset in Fig. 7 highlights
the partial pressures of the benzene groups (m / z = 77, 78), NO,
and the admolecule (m / z = 123), which are likely products of a
nitrophenyl desorption process. In this case, the species' pre-
Intensity (arb. units)
1
750C before max temp.
1089C max temp.
750C after max temp.
0.8
0.6
0.4
sence in the vacuum chamber diminished throughout the heating
cycle. Even when the temperature increased to a maximum
(1089 °C), the partial pressure did not exceed its initial value at
750°C. One explanation for this behavior is that the partial
pressure of benzene groups had peaked due to desorption from
the surface before the TEED at maximum temperature and
corresponding residual gas partial pressures were measured (the
time required to increase temperature from 750 °C to the maximum temperature was approximately 30 min). Then, when the
measurements were carried out at maximum temperature, the
partial pressure of the benzene groups had decreased because the
supply from the sample emitter's surface had already been
exhausted.
5.4. Nanocrystalline diamond with re-generated hydrogen
termination
After recording the results for the HND sample described in
Section 5.2, hydrogen termination was regenerated in a PECVD
system, and the sample's thermionic performance was reassessed. The hydrogen plasma treatment was performed at
800W, 200 sccm H2, 30 Torr and approximately 780 °C for
10min. Subsequently, the re-generated nanocrystalline diamond
(R-HND) film was slowly cooled in the presence of atomic
Table 2
Resolution, energy of peak intensity and effective work function of TEEDs from
the NND sample
0.2
0
3.5
4
4.5
5
5.5
6
Energy (eV)
Fig. 6. Sequential TEEDs from nanocrystalline diamond film with nitrophenyl
termination measured at 750, 1089 and 750°C. The shift in the energy at
maximum intensity before and after the maximum temperature indicates that the
relative area of higher work function has increased in size.
TEED
Before max.
temp.
Max. temp.
After max.
temp.
Temperature
(°C)
Energy at max.
intensity (eV)
Corresponding estimated
work function, ϕ (eV)
750
3.96
3.88
1089
750
4.66
4.63
4.57
4.55
V.S. Robinson et al. / Diamond & Related Materials 15 (2006) 1601–1608
Partial Pressure, Nitrogen Equivalent
(x10-6 torr)
2.00
1.80
1.60
1.40
750C before max temp
1089C max temp
750C after max temp
1.20
1.00
Partial Pressure, Nitrogen
Equivalent (x10-6 torr)
1606
1.20E-02
1.00E-02
8.00E-03
6.00E-03
4.00E-03
2.00E-03
0.00E+00
30
0.80
77
78
123
(m/z) Charge-to-Mass Ratio
0.60
0.40
0.20
0.00
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49
Mass-to-Charge Ratio
Fig. 7. Mass spectra from nanocrystalline diamond with a nitrophenyl adlayer at 750, 1089 and 750°C. The inset displays the partial pressure of four particular massto-charge ratios (30 and 77, 78 and 123) that correspond to NO, benzene (77 and 78), and the molecular ion of the admolecule, respectively. These species are likely
present during desorption of the nitrophenyl termination layer.
hydrogen as discussed vide supra. After restoring the surface
termination, TEEDs were measured during another heating
cycle up to a maximum temperature of 940°C.
Similarly to the previous samples, heating the R-HND appeared to cause desorption of the surface termination layer. Fig. 8
contains TEEDs measured at 700, 750 and 850°C. For temperatures at and below 750°C, thermionic emission was stable, with
repeatable TEEDs. However, at 850°C, the location of peak
intensity increased by approximately 0.9 to 4.8eV. The effective
work function of the R-HND sample below 750°C was approximately 3.9eVas determined by curve-fitting Eq. (4) assuming a
single work function. With 95% confidence, the root mean
square errors in the curve-fits for all data in Fig. 8 were less than
0.026, and the R-square values were greater than 0.965.
Fig. 9 re-plots the 750°C R-HND data and the corresponding
curve fit and also illustrates that the shift in work function with
increasing temperature is an irreversible process. The curve-fit for
Fig. 8. . TEEDs from R-HND at 700, 750 and 850°C. Distributions were
consistent below 750°C, and again after 850°C. The shift in energy of the
distribution is attributed to desorption of hydrogen from the surface.
the TEED at 750°C is representative of the excellent agreement
between theoretical and experimental distributions. In this particular case, the root mean square error was 0.004, and the R-square
value was 0.999. The TEED at 900°C in Fig. 9 is consistent with
that at 850°C in Fig. 8. After dwelling at a maximum temperature
of 940°C, the sample temperature was decreased with TEEDs
recorded at 800, 750, and 700°C. Once the distribution had
shifted to higher energy as shown in Fig. 8, the distributions were
very consistent at all temperatures and exhibited work functions
between 4.81 and 4.85eV. Each of these distributions was analyzed with the curve-fitting procedure described above, and Table
3 lists the most relevant parameters of each. These results for the
R-HND sample indicate a permanent work function increase of
approximately 0.9eV.
Comparison of the TEED shapes for the three tested samples
(HND, NND, R-HND corresponding to Figs. 4, 6, and 9),
Fig. 9. The shift to higher energies was permanent, as demonstrated by the
constant distribution with varying temperature. The TEEDs at 800, 750, and
700°C were measured after the sample had been heated to a maximum
temperature of 940°C. A representative curve fit at 750 °C is included. Curve fits
were used to estimate the work function of the R-HND sample.
V.S. Robinson et al. / Diamond & Related Materials 15 (2006) 1601–1608
Table 3
Summary of curve-fitting results for the regenerated HND (R-HND) before and
after maximum temperature
Temperature (°C)
Estimated work
function (eV)
Analyzer
resolution, σ (eV)
R2
750, before max temp
900, before max temp
800, after max temp
750, after max temp
700, after max temp
3.86
4.85
4.81
4.85
4.81
0.030
0.067
0.052
0.052
0.030
0.999
0.997
0.998
0.989
0.964
Analyzer resolution σ is defined in Eq. (3). R2 represents the proportion of
variation explained by the model [35] and varies between 0 and 1, with a value
of unity representing an ideal model.
respectively) suggests that, for the former two samples, a mixture of regions of different work functions exists across the
surfaces, as indicated by the shoulders in the distributions.
Heterogeneities in the electrical and electrochemical properties
across the surface of polycrystalline, boron-doped diamond
films are well known and have recently been investigated by
conducting probe atomic force microscopy (CP-AFM) and
scanning electrochemical microscopy (SECM) [34]. However,
the distributions from the re-generated hydrogen-terminated
surface conform well to a single-work-function model.
This result suggests that a change occurred in the chemical
and or electronic properties of the surface after rehydrogenation.
One possible explanation is the loss of much of the π-bonded
grain boundary carbon due to hydrogen chemisorption. The
initial nanocrystalline diamond film has at least two general
sites for emission: the hydrogen terminated sp3-bonded carbon
of the diamond grains and the sp2-bonded carbon in the grain
boundaries. Extended phases of sp2-bonded carbon do not exist
in the grain boundaries. The latter sites are expected to have a
higher work function than the former. After the hydrogen
plasma treatment, it is possible that much of the sp2-bonded
grain boundary carbon has been transformed to a sp3 bonding
configuration. The possibility of some etching exists, although
the morphology of the films is unaltered after plasma treatment,
but the primary modification is hydrogen chemisorption, leading to a surface with a more homogeneous work function.
Another possible explanation is that the electrical conductivity
across the surface is more uniform after rehydrogenation. Recent CP-AFM measurements support this latter supposition as it
was observed that the electrical conductivity is more uniform
over a boron-doped nanocrystalline diamond surface after
hydrogen plasma treatment.
6. Conclusions
The present work indicates that hydrogen and nitrophenyl
termination layers on the surface of boron-doped nanocrystalline
diamond become unstable at temperatures near 750°C. Comparisons of TEEDs at approximately 750°C before and after
heating to a higher temperature reveal a large, upward shift in
effective work function. Partial pressure measurements for the
hydrogen-terminated surface indicate an increase in molecular
hydrogen after heating to the maximum temperature, but similar
experiments on species associated with the nitrophenyl did not
1607
exhibit such an increase. In all cases, the increase in work function
occurred near 750°C, but the shapes of the distributions differed
in that those from the original hydrogen- and nitrophenyl-terminated samples showed shoulders that were indicative of multiple
work functions while the re-generated hydrogen-terminated surface produced distributions corresponding to a single work function. Consistent with this are (i) the variations we have recently
detected in the apparent work function across the surface of
boron-doped diamond films using scanning Kelvin-probe force
microscopy and (ii) the variations in electrical conductivity across
the surface of a boron-doped diamond film as detected by conductivity-probe atomic force microscopy [34]. The variability in
the work function deserves further examination, as uniformity of
emission is an important consideration in the application of thermionic emission materials in practical devices.
Acknowledgements
The authors gratefully acknowledge financial support for this
work through the National Science Foundation (NIRT
0210336), and Matt Maschmann's assistance in producing the
SEM images.
References
[1] G.P. Smestad, Sol. Energy Mater. Sol. Cells 82 (2004) 227.
[2] G. Miskolcy, D.P. Lieb, Proc. Intersoc. Energy Conver. Eng. Conf., Inst.
Electric. Electron. Eng., Reno, Nevada, , 1990, p. 222.
[3] H. Oman, IEEE Aerosp. Electron. Syst. Mag. 18 (2003) 28.
[4] V.I. Yarygin, Ye.A. Meleta, S.M. Tulin, V.V. Klepikov, V.A. Ruzhnikov,
L.R. Wolff, Proc. Proc. Intersoc. Energy Conver. Eng. Conf., Inst. Electric. Electron. Eng., Monterey, California, , 1994, p. 1061.
[5] T.H. Van Hagan, J.N. Smith Jr., Proc. Intersoc. Energy Conver. Eng. Conf.,
Inst. Electric. Electron. Eng., Washington, D.C., 1996, p. 629.
[6] S.W. Angrist, Direct Energy Conversion, Allyn and Bacon, Inc., Boston,
1965.
[7] K.L. Thayer, M.L. Ramalingam, T.R. Lamp, Proc. Int. Mech. Eng. Conf.
Expo., Am. Soc. Mech. Eng., Chicago, IL., 1994, p. 1.
[8] T.S. Fisher, Appl. Phys. Lett. 79 (2001) 3699.
[9] T.S. Fisher, D.G. Walker, J. Heat Transfer 124 (2002) 954.
[10] V.S. Robinson, T.S. Fisher, J.A. Michel, C.M. Lukehart, Appl. Phys. Lett.
87 (2005) 061501.
[11] F.A.M. Köck, J.M. Garguilo, B. Brown, R.J. Nemanich, Diamond Relat.
Mater. 11 (2002) 774.
[12] J.B. Cui, J. Ristein, L. Ley, Diamond Relat. Mater. 9 (2000) 1143.
[13] L. Diederich, O.M. Küttel, P. Aebi, L. Sclapbach, Diamond Relat. Mater.
8 (1999) 743.
[14] T.-C. Kuo, R.L. McCreery, G.M. Swain, Electrochem. Solid-State Lett. 6
(1999) 288.
[15] J. Pinson, F. Podvorica, Chem. Soc. Rev. 34 (2005) 429.
[16] J.W. Gadzuk, E.W. Plummer, Rev. Mod. Phys. 45 (1973) 487.
[17] R.D. Young, Phys. Rev. 113 (1959) 110.
[18] A.C. Marshall, Surf. Sci. 517 (2002) 186.
[19] R.D. Young, C.E. Kuyatt, Rev. Sci. Instrum. 39 (1968) 1477.
[20] R. Reifenberger, H.A. Goldberg, M.J.G. Lee, Surf. Sci. 83 (1978) 599.
[21] I. Brodie, P.R. Scwoebel, Proc. IEEE 87 (1994) 1006.
[22] G. Popovici, M.A. Prelas, Phys. Status Solidi, A 132 (1992) 233.
[23] M. Delamar, R. Hitmi, J. Pinson, et al., J. Am. Chem. Soc. 114 (1992)
5883.
[24] A.O. Solak, L.R. Eichorst, W.J. Clark, R.L. McCreery, Anal. Chem. 75
(2003) 296.
[25] J. Wang, M.A. Firestone, O. Auciello, J. Carlisle, Langmuir 20 (2004)
11450.
1608
[26]
[27]
[28]
[29]
[30]
[31]
[32]
V.S. Robinson et al. / Diamond & Related Materials 15 (2006) 1601–1608
M.A. Brown, L. Neelands, H. Farnsworth, J. Appl. Phys. 21 (1950) 1.
M. Nichols, Phys. Rev. 57 (1940) 297.
G. Smith, Phys. Rev. 94 (1954) 295.
R.D. Young, Phys. Rev. 113 (1959) 110.
R.C. West, CRC Handbook of Chemistry and Physics, CRC Press, 1976.
D.M. Gruen, Annu. Rev. Mater. Sci. 29 (1999) 211.
J. Birrell, J.E. Gerbi, O. Auciello, J.M. Gibson, D.M. Gruen, J.A. Carlisle,
J. Appl. Phys. 93 (2003) 5606.
[33] A.V. Hamza, G.D. Kubiak, R.H. Stulen, Surf. Sci. 237 (1990) 35.
[34] K.B. Holt, A.J. Bard, Y. Show, G.M. Swain, J. Phys. Chem., B 108 (2004)
15117.
[35] N.R. Draper, H. Smith, Applied Regression Analysis, Third Edition, John
Wiley and Sons, Inc., New York NY, 1998.